[Bug libgcj/10353] [3.4/4.0/4.1/4.2 regression] Java testsuite failures

["ebotcazou at gcc dot gnu dot org" <gcc-bugzilla@gcc.gnu.org>]

------- Comment #25 from ebotcazou at gcc dot gnu dot org  2006-01-29 17:07 -------
Status for 4.1.0pre on Solaris 7, 8, 9 and 10:

32-bit:

                === libjava tests ===


Running target unix
FAIL: Array_3 execution - gij test
FAIL: Array_3 execution - gij test
FAIL: FileHandleGcTest execution - source compiled test
FAIL: FileHandleGcTest execution - gij test
FAIL: FileHandleGcTest execution - bytecode->native test
FAIL: FileHandleGcTest -O3 execution - source compiled test
FAIL: FileHandleGcTest execution - gij test
FAIL: FileHandleGcTest -O3 execution - bytecode->native test

                === libjava Summary ===

# of expected passes            3995
# of unexpected failures        8
# of expected failures          10
# of untested testcases         16


64-bit:

                === libjava tests ===


Running target unix
FAIL: Array_3 execution - gij test
FAIL: Array_3 execution - gij test
FAIL: FileHandleGcTest execution - source compiled test
FAIL: FileHandleGcTest execution - gij test
FAIL: FileHandleGcTest execution - bytecode->native test
FAIL: FileHandleGcTest -O3 execution - source compiled test
FAIL: FileHandleGcTest execution - gij test
FAIL: FileHandleGcTest -O3 execution - bytecode->native test
FAIL: Throw_2 execution - gij test
FAIL: Throw_2 execution - gij test

                === libjava Summary ===

# of expected passes            3991
# of unexpected failures        10
# of expected failures          10
# of untested testcases         18


i.e. we are on par with GCC 4.0.x on both 32-bit and 64-bit SPARC/Solaris.

There is a significant difference with GCC 4.0.x though: the above results are
valid only with GNU binutils (in this case 2.16.1 CVS), not with the Sun tools;
in the latter case, there are several additional failures:
  http://gcc.gnu.org/ml/gcc-testresults/2006-01/msg01501.html
  http://gcc.gnu.org/ml/gcc-testresults/2006-01/msg01499.html

I've investigated a little and the additional failures seem to be related to
strings and UTF-8 support.  I don't really plan to dig further.  I'll add a
blurb about that in the platform-specific installation notes.


-- 

ebotcazou at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|3.3.3 3.4.0 4.0.0           |3.3.3 3.4.0 4.0.0 4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10353